ENGINE Error in 'start' listener

[ncjok]

I've come across a resource hogging problem with sabnzbdplus 0.5.4-1 on a remote (internet) server running Debian squeeze. Here is some hopefully relevant console output: http://pastebin.com/raw.php?i=eHhcAdqg

One minute after start up, sabnzbdplus starts to hog a full CPU core repeatedly for periods of about 20 sec on, 10 sec off. The cherrypy errors seem to suggest a problem binding the desired web interface port, however lsof -i and netstat show a listening socket. Additionally, the web interface remains accessible following the final "All processes stopped" log entry.

In /etc/default/sabnzbdplus, the USER variable is employed to drop privileges. The variable HOST has been tried empty and with 0.0.0.0, while for PORT I have tried 8080, 8181 and 16362. All combinations lead to the undesirable behaviour noted above.

I hope someone can decipher any applicable errors and help me get SABnzbd+ running nicely

[jcfp]

Are you sure the program isn't already running (one way or the other)? Relevant output of the ps command before/during/after the attempt to start the service?

[ncjok]

Hi jcfp,
Thanks for your reply. I was pretty sure that there wasn't a problem with duplicated processes or something else having already bound a socket with the port I wanted SABnzbd+ to use. Your comment did, however, prompt me to try and keep a closer eye on the process during start up and through the first minute to the point at which it starts to get out of control. To that end, I had resorted to adjusting PORT in /etc/default/sabnzbdplus again and rebooting the system, following which I caught some additional communication to the port in question,

Code: Select all

root@135:~# lsof -i | grep sab
sabnzbdpl 870    share    6u  IPv4   3974      0t0  TCP *:8181 (LISTEN)
sabnzbdpl 870    share    8u  IPv4   4080      0t0  TCP localhost:42625->localhost:8181 (SYN_SENT)

Now I should apologise for not disclosing the fact I had tried to implement firewall (iptables) rules to restrict web interface access to my home IP only.

Code: Select all

root@135:~# iptables -L -v -n (snipped to show relevant output)
Chain INPUT (policy ACCEPT 78 packets, 7390 bytes)
 pkts bytes target     prot opt in     out     source               destination
 1485  111K ACCEPT     all  --  *      *       0.0.0.0/0            0.0.0.0/0           state RELATED,ESTABLISHED
   0     0 ACCEPT     tcp  --  *      *       MyHomeIP             0.0.0.0/0           tcp dpt:8181 state NEW
 150 12835 DROPIN     all  --  *      *       0.0.0.0/0            0.0.0.0/0

Chain DROPIN (1 references)
 pkts bytes target     prot opt in     out     source               destination
  53  3180 DROP       tcp  --  *      *       0.0.0.0/0            0.0.0.0/0           tcp dpt:8181

The input rules were therefore adjusted to ignore the loopback interface,

Code: Select all

root@135:~# iptables -L -n -v (snipped)
Chain INPUT (policy ACCEPT 0 packets, 0 bytes)
 pkts bytes target     prot opt in     out     source               destination
  56  4196 ACCEPT     all  --  !lo    *       0.0.0.0/0            0.0.0.0/0           state RELATED,ESTABLISHED
   0     0 ACCEPT     tcp  --  !lo    *       MyHomeIP             0.0.0.0/0           tcp dpt:8181 state NEW
   0     0 DROPIN     all  --  !lo    *       0.0.0.0/0            0.0.0.0/0

Chain DROPIN (1 references)
 pkts bytes target     prot opt in     out     source               destination
   0     0 DROP       tcp  --  *      *       0.0.0.0/0            0.0.0.0/0           tcp dpt:8181

...and this appears to have solved the problem

[ncjok]

Indeed. Thanks for that useful tidbit of information, it has the potential to solve some of my future oversights

[dilibero]

thanks for these post, it solve my problem